Data center operators sometimes find it hard to get a chance to pitch to new business. So much so that they are considering extreme measures. A wholesale colocation operator admitted to a colleague at DCD how hard it is to approach hyperscale customers. There’s a limited number of them, and the decision-makers within those companies can be difficult to access.
Our contact then confided that his company had seriously considered using private detectives to track down the actual people involved in buying colocation! Armed with his prospects’ personal details, he would be able to advertise on billboards they would pass on their commute. He could sponsor the little leagues their children played in. He would engineer ways to prime them subliminally with his company’s name.
This wasn’t a fantasy. It was an actual strategy he had considered. He eventually rejected it, and we’re not sure why. Perhaps, after savouring a few gumshoe cliches from film noir, he realised this was not an attractive and romantic prospect. In truth, it would just make his company look creepy. Wait up - it would make his company ACTUALLY creepy.
I suspect his team evaluated the idea and found it wasn’t cost-effective. They probably actually found that in-house staff could probably do just as good a job much more cheaply, using social media. For whatever reason, he passed on the concept. But he did assure us that he knew other operators which are actually doing this form of prospect identification and tracking.
Maybe he is right. Customer acquisition is a serious issue. In wholesale colocation, there are a limited number of hyperscale customers buying in large quantity, so that’s a small list of prospects, and you need to be sure they all know your name.
Retail colocation and the enterprise market is a different thing. There are more potential customers, they buy in smaller amounts, and they can come and go, so more normal rules apply. Most wholesale colocation operators serve a mixture of those two markets, but they can turn the dial a long way in either direction.
Many projections of the progress of cloud suggest that colocation is all moving in the direction of hyperscale. The story tells that enterprise customers are using a form of “hybrid cloud”, combining of in-house IT with a growing slice of cheap, on-demand services.
I found a lot of operators shifting towards hyperscale customers on a recent tour of Ashburn in Northern Virginia, also known as “Data Center Alley”. By a long chalk, this is the greatest concentration of data centers in the world. And all the operators I spoke to said their business is slanted towards giant hyperscale customers. The balance of their business is shifting to 60 or 70 percent hyperscale, or more.
Our contact has gone even further. He only deals with hyperscale giants.
He probably believes he is ahead of the game. But in his world margins are very tight. Hyperscalers want capacity quickly and they want it at the lowest possible cost per MW, but with conventional levels of reliability.
In that world, film noir-style customer acquisition might be an interesting idea to kick around, but, like everything else except basic service provision, it will ultimately be unaffordable as the customers crush the margins.
Other sectors, like enterprise colocation, retail colocation and specialist services like high performance computing (HPC) will, as always, fall back on conventional marketing methods. So forget the image of a gumshoe lurking on Waxpool Road. For most of us in the real world, fishing for prospects that way is wrong. And for those who might feel forced to look at it, it is probably too expensive.